Pseudo-Consciousness in Artificial Intelligence: a functional and governance framework for consciousness-like systems
PhilSci-Archive April 5, 2026 José Augusto de Lima Prestes
Advanced AI systems that perform many functions associated with consciousness—integrating information, revising outputs, transferring skills across domains, simulating goals, and maintaining a consistent behavioral profile—should be understood through a new category called 'pseudo-consciousness,' rather than being attributed inner subjective experience. This concept occupies the middle ground between reactive automation and genuinely conscious agents. Pseudo-consciousness provides a disciplined vocabulary for systems whose social effects exceed older categories like narrow AI, and clarifies why they generate distinctive ethical and governance concerns even without evidence for consciousness. The article develops five conditions for identifying pseudo-conscious profiles and examines boundary cases involving large language models, multimodal systems, and tool-using agents.